The Efficacy of a Single-assignment Writing Therapy on Posttraumatic Stress Symptoms Yvonne Vinke
Arnold van Emmerik 5697 words Masterthesis 234 words abstract Date: 25-07-2016
1 Abstract
This study examined whether a single-assignment writing therapy is effective in decreasing posttraumatic stress (PTS) symptoms and depressive symptoms. Furthermore, this study examined whether this effect was moderated by alexithymia, the locus of control or posttraumatic cognitions. 49 participants were divided over three conditions: in the first condition participants had to write about a traumatic event, in the second condition participants had to write about neutral images and in the last condition participants were placed on a waiting list. Directly after the screening,
participants received questionnaires measuring depressive symptoms, alexithymia, locus of control and posttraumatic stress cognitions. One week later, the participants of two conditions had to make a writing assignment, either about the traumatic event or about neutral images. One week and one month after this assignment or, for the waiting list condition, two and five weeks after the first questionnaires, PTS- and depressive symptoms were measured again. Unexpectedly, results showed that PTS symptoms decreased over time, regardless of the condition. Furthermore, depressive symptoms only decreased over time in the condition that had to write about neutral images. Finally, there was no moderating effect of alexithymia, posttraumatic cognitions or locus of control. Possibly, the disclosure in the pre-test overrules the effect of one trauma-focused writing assignment on the PTS symptoms. A future study might focus on the contribution of multiple writing assignments to this effect of the pre-test on the PTS- and depressive symptoms.
2 Table of contents
Introduction ...3
Methods ...7
Participants ...7
Measures for eligibility ...8
Psychotic symptoms ...8
Suicidal ideation ...8
Posttraumatic stress ...9
Depressive symptoms ...9
Measures for moderator variables ... 10
Posttraumatic cognitions ... 10 Alexithymia ... 10 Locus of control ... 11 Procedure ... 11 Statistical analyses ... 12 Results ... 13 Baseline equivalence... 13
Change in PDS-scores and BDI-II scores ... 14
Moderation analyses ... 16
Discussion... 17
3 Introduction
Almost everyone experiences daily hassles. Even though a serious trauma is rare, a survey study of Resick (2000) showed that 69% of the surveyed persons aged 18 to 60+ experienced at least one traumatic stressor in their life and 21% experienced a traumatic stressor in the past year. Resick (2000) defined a traumatic stressor as “(…) those events beyond daily hassles, beyond normal developmental life challenges, beyond more stressful and challenging circumstances such as divorce, losing a job, serious illness, or financial problems. (…) events that are life threatening (…) and that are accompanied by intense fear, helplessness, or horror (p. 2)”. These traumas can lead to
posttraumatic stress (PTS) symptoms such as: persistent re-experiencing of the traumatic event (intrusive recollection), avoidance of stimuli associated with the traumatic event, numbing of general responsiveness and hyper-arousal. Approximately 7% will develop posttraumatic stress disorder (PTSD; American Psychiatric Association, 2000).
The American Psychiatric Association (APA; 2000) recommends multiple face-to-face psychotherapeutic interventions for persons who develop PTSD to reduce the PTS symptoms, to prevent/treat comorbid conditions like depression and to protect someone against relapse. These treatments, like the recommended eye movement desensitization reprocessing (EMDR), mostly involve exposure to the traumatic event (Hendriks, de Kleine, & van Minnen, 2015). However, the drop-out rate for treatments like EMDR is 20-35%, possibly because it requires frequent face-to-face contact between the therapist and client (Schnurr et al., 2015). Kuester, Niemeijer and Knavelsrud (2016) noted that “(…) around one in five patients seeks psychological treatment due to fear of stigmatization, embarrassment, judgement or exclusion and, or negative beliefs about mental health care services (p. 3)”. Another reason for this drop-out possibly is that it requires multiple exposures to the traumatic event before a clinically significant decrease in symptoms is visible. The anxiety during each exposure without the relieve between the therapy sessions could possibly cause persons to drop-out of treatment. Therefore, a treatment which alleviates PTS symptoms immediately might decrease the drop-out rate.
4 Besides face-to-face therapies, the number of empirically supported e-mental health
interventions like Interapy (Lange et al., 2000) is growing. Another empirically supported e-mental health intervention is the online writing therapy for PST; cognitive behavioral writing therapy (CBWT; van Emmerik, Reijntjes, & Kamphuis, 2013). CBWT consists of the following treatment elements: psycho-education, exposure, cognitive restructuring, promoting adequate coping and social sharing (Van der Oord, Lucassen, van Emmerik, & Emmelkamp, 2010). Since this writing therapy is
administered via the Internet, it is ideal for persons who live in remote areas or who want to stay anonymous due to, for example, fear of stigmatization. Preliminary evidence suggested that the effect of a writing therapy on PTS symptoms not only remain on the short-, but also on the long-term (van Emmerik, Kamphuis, & Emmelkamp, 2008).
Results from Truijens and van Emmerik (2014) suggest that even a single writing assignment might be efficacious in immediately relieving PTS symptoms. If a single-assignment writing therapy would be efficacious, it would not only help patients by reducing their symptoms quickly, it would also reduce treatment costs. Therefore, it is necessary to further examine whether a single-assignment writing therapy, as opposed to multiple single-assignment writing therapy, is effective in decreasing PTS symptoms.
A single-assignment writing therapy also looks promising when examining the current
‘acknowledged’ theory of PTSD; the cognitive theory of Ehlers and Clark (2000). This theory suggests that persons with PTSD have a sense of serious, current threat resulting from: (1) excessive negative appraisals regarding the trauma and/or its sequelae and (2) a disconnection between the traumatic memory and the autobiographical memory. The PTSD symptoms result from the current sense of threat. Finally, this theory claims that persons with PTSD use cognitive and behavioral avoidance strategies to reduce their sense of threat, which ironically leads to the maintenance of the sense of threat. Ehlers and Clark (2000) propose that three areas require change when treating PTSD: (a) the traumatic memory must be integrated into the autobiographical memory, (b) appraisals of the
5 traumatic event and/or its sequelae must be modified and (c) one must abandon cognitive and behavioral avoidance strategies since these hinder the first two processes.
A single-assignment writing therapy could have an effect on the three areas requiring change when treating PTSD: (a) the exposure might promote the connection between the unconnected traumatic memory and the autobiographical memory. (b) Writing about the trauma could facilitate the retrieval of traumatic memories difficult to retrieve otherwise, since the written exposure requires someone to try to re-experience the traumatic event as fully as possible in their mind. These new traumatic memories might lead to reappraisal of the traumatic event; for example: when someone learns he/she did scream for help during the event it could diminish the appraisal that he/she deserves it that bad things happen to him/her. (c) To be able to write about the traumatic event, one has to abandon strategies to avoid thinking about the traumatic event. Therefore, a single-assignment writing therapy might be effective in reducing PTS symptoms.
Since depressive symptoms are most commonly comorbid with PTS symptoms (Stander, Thomsen, & Highfill-McRoy, 2014), and since a writing therapy seems effective in decreasing these comorbid depressive symptoms (van Emmerik, Reijntjes, & Kamphuis, 2013),a single-assignment writing therapy might also alleviate depressive symptoms. The current study will examine whether a single-assignment writing therapy, opposed to a multiple assignment writing therapy, is able to alleviate PTS- and depressive symptoms.
When examining the effectiveness of a single-assignment writing therapy, possible moderating factors influencing the effect of the therapy on the PTS- and depressive symptoms should be considered. In writing therapy, one factor could be the locus of control; the ways in which individuals attribute environmental events. An internal locus of control indicates that an individual attributes events to himself, while someone with an external locus of control attributes events to things beyond his power (Rotter, 1966). Since an internal locus of control is associated with the effort to improve functioning (Strickland, 1978), it could be expected that someone who has more of an external outlook (external locus of control) would benefit less from writing therapy since they might
6 expose themselves less. In other words, it is expected that the locus of control moderates the effect of a single-assignment writing therapy: a more internal outlook could possibly lead to more
exposure, resulting in a higher benefit from the single-assignment writing therapy.
Another moderating factor that might influence the outcome of writing therapy on PTS symptoms is alexithymia. According to Parker, Eastabrook, Keefer and Wood (2010) “alexithymia is associated with (…) difficulty identifying and describing feelings, difficulty distinguishing between feelings and bodily sensations, constricted imaginal processes, and a cognitive style that is concrete and externally focused” (p. 798). It is hypothesized that persons with more alexithymic
traits/characteristics benefit less from writing therapy, since their inability to identify their feelings might hinder with the process of fully re-experiencing the trauma. Therefore, the integration of the traumatic memory in the autobiographical memory might be more difficult. Another reason that a person with more alexithymic traits might benefit less from writing therapy is because of its correlation with more severe PTS symptoms (Frewen et al., 2008).
Finally, dysfunctional PTS cognitions might have a moderating effect on a single-assignment writing therapy. Persons with PTS symptoms often experience an increase in dysfunctional cognitions about themselves, others and the world (Dunmore, Clark, & Ehlers, 1999). A reduction in these cognitions is related to a reduction in PTS symptoms (van Emmerik, Schoorl, Emmelkamp, & Kamphuis, 2006). Since a single-assignment writing therapy does not include extensive cognitive restructuring, it is highly likely that very firm dysfunctional cognitions decrease the effect of the single-assignment writing therapy on the PTS symptoms. The anxiety (or other negative emotions) caused by re-experiencing the trauma due to the writing assignment would probably activate these very firm dysfunctional cognitions, thereby interfering with the treatment.
The present study investigated the following two hypotheses concerning the effectiveness of a single-assignment writing therapy: (1a-1b) a single writing assignment focusing on the trauma is effective in reducing PTS- and depressive symptoms and (2a-2c) alexithymia, the locus of control and posttraumatic cognitions moderate this effect. It was expected that participants with a more external
7 outlook, who have more alexithymic traits or who have more dysfunctional posttraumatic cognitions benefit less from the single-assignment writing therapy. In view of the empirically supported online writing therapy for PTS, the study was conducted via a secured and reliable server space in Qualtrics (Ruwaard, Lange, Schrieken, Dolan, & Emmelkamp, 2012).
Methods Participants
After approval of the Commission of Ethics, information about the study was disseminated through DPMS and information leaflets distributed in the main buildings of the University of Amsterdam (UvA). DPMS is a website on which potential participants sign up if they wish to participate in a study carried out at the UvA. The objective was to recruit 451 participants in total, resulting in 15 participants in each condition. This way sufficient statistical power would be retained despite a probable drop-out of three per condition.
Eligibility criteria included (a) sufficient fluency in Dutch to complete the study procedures, (b) clinically elevated PTS symptoms as evidenced by a score of 18 or higher on the Posttraumatic Diagnostic Scale, (c) an absence of psychotic symptoms as evidenced by a score below 5 on the Screening Device for Psychotic Disorder, and (d) absence of current suicidal ideation as evidenced by a score of one on a 2-point Likert scale item on the Suicide Risk List, and a score of 0 on the Beck Depression Inventory – II suicidality item. After it was determined via a screening that participants met eligibility criteria, of 197 applicants, 49 (25%) were divided over the Writing about Trauma (WAT; n = 17) condition, Writing about something Trivial (WT; n = 18) condition and the Waiting List (WL; n = 14) condition. Participant 1 was assigned to condition 1, participant 2 to condition 2, participant 3 to condition 3, participant 4 to condition 1 etcetera.
Participants who completed the study either received participation credit points with a ratio of one per hour or a financial remuneration with a maximum of €17.50.
1 Power analysis show that at least 36 participants in total (12 per condition) are necessary for enough power. Based on earlier research, the chosen effect size is f = 0.25 (Truijens & van Emmerik, 2014). The chosen α = 0.05, power = 0.8, number of groups = 3 and number of measurements = 3.
8 Measures for eligibility
Psychotic symptoms
The Dutch Screening Device for Psychotic Disorder (SDPD) has been used to screen psychotic symptoms (Lange, Schrieken, Blankers, van de Ven, & Slot, 2000). The SDPD is an eight-item self-report questionnaire including two subscales: one measuring general psychotic symptoms (six items) and one measuring auditory hallucinations (two items). Participants have to indicate to what extent certain psychotic symptoms have been applicable to them in the past five years. The items are rated on a 5-point Likert Scale (1 = not at all, 5 = very often). The overall score on the questionnaire ranges between 8 and 40, with higher scores indicating more psychotic symptoms. Two examples of items are: “In the past five years it has occurred that I heard voices other persons could not hear” and “In the past five years it has occurred that I saw things other persons could not see’. The internal consistency of the SDPD is good, with a Cronbach’s alpha value of .73 (Lange et al., 2000).
Participants were excluded if their overall SDPD score was 5 or higher, the cut-off score for a general psychosis (Lange et al., 2000).
Suicidal ideation
To screen for suicidal ideation, the Dutch version of the Suicidal Risk List has been used (SRL; Lange et al., 2005, adapted from Blaauw, Voort, & Kerkhof, 1999). The SRL is a six-item self-report questionnaire evaluating current feelings of desperateness and suicidal intentions and the history (time and location) of previous suicide attempts. The items are rated on a 2-point (1 = no, 2 = yes) or 3-point Likert Scale (1 = no, 2 = once, 3 = more than once). Higher scores indicate more suicidal ideation. Two examples of items are: “Have you ever tried to take your life?” and “Are you currently preparing to take your life?”. Potential participants with a score of 2 on a 2-point Likert Scale would be excluded since this indicates acute suicidal intentions.
9 Posttraumatic stress
Posttraumatic symptom severity and functioning has been assessed with the Posttraumatic Diagnostic Scale (PDS; Foa, Cashman, Jaycox, & Perry, 1997). The PDS is a 49-item self-report measure with four sections. Part one is a traumatic event checklist on which participants checkmark events they have witnessed or experienced. In part two participants are asked which of the
checkmarked events bothers them the most and to briefly describe this event, also determining whether DSM-IV-TR criteria A1 and A2 are met (American Psychiatric Association, 2000). The third part consists of rating which of the 17 cardinal symptoms of PTSD a participant experienced in the past 30 days. These 17 items are rated on a 4-point Likert scale (0 = not at all, 3 = very much) over a period of the past month. Finally, part four ascertains the level of impairment in certain life
functioning areas. The overall PDS score ranges from 0 to 51 and is based on the scores obtained in part three. Cutoff scores for symptom severity rating are: 0 = no rating, 1-10 = mild, 11-20 =
moderate, 21-35 = moderate to severe and 36> severe PTS symptoms. Two examples of items are: “Having bad dreams or nightmares about the traumatic event” and “Feeling irritable or having fits of anger” (Foa, Cashman, Jaycox, & Perry, 1997). Test-retest reliability using symptom severity scores is high (r = .83; McCarthy, 2008). Potential participants were excluded if their overall PDS score was below 19, a cut-off score yielding perfect sensitivity and adequate specificity (Wohlfarth, Van den Brink, Winkel, & Ter Smitten, 2003). Besides using the PDS to screen participants, this study used the PDS to indicate whether the trauma-focused writing assignment had an effect on the PTS symptoms.
Depressive symptoms
The Beck Depression Inventory-II (BDI-II) has been used to measure depressive symptoms (Beck, Steer, & Brown, 1996). The BDI-II is a 21-item self-report scale. Items endorsed statements characterizing how a participant was feeling throughout the past two weeks. Items were rated on a 4-point Likert scale ranging from 0 to 3. The overall score on the BDI-II ranges from 0 to 63, with higher scores indicating more depressive symptoms. Overall scores of 0 to 13 denote minimal
10 depression, scores of 14 to 19 denote mild depression, scores of 20-28 denote moderate depression, and scores of 29-63 denote severe depression (Wang & Gorenstein, 2013). An example of an item is: “Sadness: (0) I do not feel sad, (1) I feel sad most of the time, (2) I feel sad all the time or (3) I feel so sad or unhappy that I can’t stand it”. With a Cronbach’s alpha of .91, the test-retest reliability in a university sample was high (Sprinkle et al., 2002). Potential participants were excluded if they indicated that they were suicidal, as evidenced by a score above 0 on the item “suicidal thoughts or wishes”. Besides using the BDI-II to screen participants, this study used the BDI-II to measure the effect of the trauma-focused writing assignment on the depressive symptoms.
Measures for moderator variables
Posttraumatic cognitions
The Dutch version of the Posttraumatic Cognitions Inventory (PTCI) has been used to measure trauma-related thoughts and beliefs (van Emmerik, Schoorl, Emmelkamp, & Kamphuis, 2006, translation of Foa, Ehlers, Clark, Tolin, & Orsillo, 1999). The PTCI is a 36-item self-report questionnaire including three subscales: negative cognitions about the self (ncs), negative cognitions about the world (ncw) and self-blame (sb). Each item is rated on a 7-point Likert scale (1 = totally disagree, 7 = totally agree). The overall score on the PTCI ranges from 36 to 252, with higher scores indicating a stronger endorsement of negative cognitions. Two examples of items are: “The world is a dangerous place” and “Nobody cares about me”. The internal consistency of the PTCI is high with an overall Cronbach’s alpha of. 93. Furthermore, the convergent validity for the PTCI is significant (r = .43, p < .001; van Emmerik, Schoorl, Emmelkamp, & Kamphuis, 2006); the PTCI correlates with PTSD symptoms.
Alexithymia
To measure alexithymia, the 20-item Toronto Alexithymia Scale (TAS-20) has been used (Taylor, Ryan, & Bagby, 1985). This self-report questionnaire measures the difficulty identifying feelings, the difficulty describing feelings, and externally oriented thinking. Participants rate their
11 endorsement with a statement on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Overall score ranges from 20 to 100, with higher scores indicating more alexithymia. “I am often confused about what emotion I am feeling” and “I prefer talking to persons about their activities rather than their feelings” are two examples of items. The TAS-20 has a good internal consistency (r = .86; Parker, Graeme, & Bagby, 2003).
Locus of control
The Internal External locus of control scale (I-E scale) has been used to measure the generalized expectancy or locus of control of a participant (Rotter, 1966). The I-E scale is a 40-item self-report scale. 26 items are rated on a 4-point to 6-point Likert scale on which participant indicate their endorsement with each statement (1 = absolutely not true, 6 = absolutely true). An example of such an item is “Most adversities happening to us are consequences of incapability, ignorance, laziness or all three”. For the last 14 items, participants have to indicate their endorsement with a statement on a visual analogue scale (VAS). The VAS is a 10 centimeters long line representing a score between 0 and 100, with the extremes representing opposite answers. An example of a statement is: “Voting during the elections is…”, with the extremes on the VAS representing “0 = very important” and “100 = useless”. The first 22 items are scored with a 1 if one of the two or three highest options is selected. The last 12 items are scores with a 1 if the participant rated the statement above a score of 50. Overall score ranges from 0 to 40, with higher scores indicating a more external locus of control. A score below 8 denotes primarily an internal locus of control and a score above 27 denotes primarily an external locus of control. Studies show that the reliability for the overall score is high, with a Cronbach’s alpha of .76 (Andriessen & van Cadsand, 1983).
Procedures
All study procedures were conducted within a secured web domain developed for the current study. Potential participants received an e-mail with an information brochure including a description of the study’s aim and procedures, a link to the website and personalized log-in codes.
12 Participants who filled in the informed consent file were randomly divided over the three conditions, after which the screening measures were administered (PDS, SRL and SDPD). Participants who were excluded received information on mental health care opportunities outside the study, tailored to their scores on these measures. Included participants completed the other pre-test measures (TAS-20, I-E scale, BDI-II, PTCI) and additional questionnaires on trauma and demographic characteristics.
Participants in the WAT and WT condition received an e-mail one week after the pre-test containing the website for the 45-minutes-long writing assignment. Participants in the WAT condition were instructed to describe their most traumatic moment in first person and present tense, and to focus on emotional and sensory experiences. The participants in the WT condition were asked to describe pictures in first person and in the present tense, and to focus on factual information.
All participants received an e-mail with the instruction to complete the PDS and BDI-II two and five weeks after the pre-test (post-test and follow-up). After the last questionnaires were completed, participants received the course credit points or the financial remuneration. Statistical analyses
The data were analyzed using IBM SPSS Statistics 20 for Windows. Pre-test equivalence between the conditions regarding demographic characteristics and overall scores on the pre-test/screening questionnaires were examined with one-way between-subjects ANOVAs and a chi-square test. To test hypothesis 1, two 2 x 3 (time by condition) repeated-measures ANOVAs were conducted with condition as the between-subjects factor and time as the within-subjects factor, to examine changes in BDI-II-scores and PDS-scores from pre-test to follow-up in the three study conditions. All statistical tests were two-tailed with an alpha level of .05. To test hypotheses 2a-2c, PROCESS (Hayes, 2013) was used to examine whether the score on the I-E scale, the PTCI and/or the TAS-20 (moderators) influenced the effect of the condition (independent variable) on the PDS-score and BDI-II-scores (dependent variables).
13 Results
Baseline equivalence
Out of 197 potential participants, the data of 44 (22%) participants could be used for data-analysis: 14 participants in the WL, 14 in the WAT and 16 in the WT condition (Figure 1). One-way ANOVAs and a chi-squared test demonstrated that the three conditions were equivalent at pre-test regarding demographic characteristics and overall scores on the questionnaires (Table 1).
Figure 1. Flow-chart showing the number of participants within each condition in the different phases of this study. DA = data-analysis; FU = follow-up; n = number of participants; PT = post-test; WT = Writing about something Trivial; WAT = Writing about Trauma; WL = Waiting List.
Potential participants (n = 197) Participants excluded (n = 145): - 72 PDS ≤ 18 - 13 SDPD ≥ 5 - 16 cancellations - 44 non-responders Included participants (n = 52) Drop-out (n = 3): - 2 cancellations - 1 non-responder WL (n = 14) WT (n = 18) WAT (n = 17) PT WL (n = 14) PT WT (n = 18) PT WAT (n = 17) Drop-out (n = 4): - 2 non-responders WT - 2 non-responders WAT FU WL (n = 14) FU WT (n = 16) FU WAT (n = 15) Participant excluded (n = 1): - 1 technical problem DA WL (n = 14) DA WT (n = 16) DA WAT (n = 14)
14
TABLE 1
Demographic Characteristics and Baseline Questionnaire Scores of Participants in the WAT, WT and WL Conditions at Pre-test
Variable WAT (n = 14) WT (n = 16) WL (n = 14) Statistic
Demographic Characteristics Females (%) 76.5% 73.7% 66.7% X2 (2) = 0.679, p = .878 Age (M ± SD) 23.24 ± 4.49 22.28 ± 2.37 24.60 ± 7.35 F (2,47) = 0.885, p = .419 Measures of eligibility BDI-II (M ± SD) 18.29 ± 6.01 18.28 ± 6.71 19.33 ± 5.65 F (2,47) = 0.151, p = .860 PDS (M ± SD) 24.18 ± 7.09 24.16 ± 5.79 23.33 ± 4.92 F (2,48) = 0.101, p = .904 SRL (M ± SD) 2.65 ± 1.00 2.79 ± 0.91 2.87 ± 1.19 F (2,48) = 0.190, p = .827 Measures for moderator variables
SDPD (M ± SD) 11.12 ± 3.24 10.26 ± 2.45 10.73 ± 2.40 F (2,48) = 0.444, p = .644 TAS-20 (M ± SD) 55.71 ± 9.22 54.17 ± 10.10 53.93 ± 8.96 F (2,47) = 0.172, p = .843 I-E (M ± SD) 16.82 ± 4.26 18.18 ± 1.88 19.27 ± 4.27 F (2,46) = 1.836, p = .171 PTCI (M ± SD) 108.59 ± 27.84 114.22 ± 33.78 110.07 ± 23.83 F (2,47) = 0.177, p = .839 Note. BDI = Beck Depression Inventory; I-E = Internal - External; SDPD = Screening Device for Psychotic Disorder; SRL = Suicide Risk List; TAS = Toronto Alexithymia Scale; PDS = Posttraumatic Diagnostic Scale; PTCI = Posttraumatic Cognitions Inventory; WAT = writing about trauma; WL = waiting list; WT = writing about something trivial.
Change in PDS-scores and BDI-II scores
PDS-scores and BDI-II scores for the three conditions from pre-test to follow-up are graphed in Figure 2 and Figure 3 respectively. Results showed a significant main effect of time, F (2, 80) = 38.182, p < .001. This means that a significant PDS-score reduction was found for all three conditions. Unexpectedly, no significant interaction effect was found between time and condition for the PDS-score, F (4, 80) = 0.630, p = .643. The same result was gathered for the BDI-II scores: there was a significant main effect of time, F (2, 80) = 5.966, p = .004. This means that a significant BDI-score
15 reduction was found for all three conditions. However, no significant interaction effect was found between time and the condition, F (4, 80) = 1.138, p = .345. This means that, regardless of the
condition someone was in, PDS-scores and BDI-II scores were always lower at post-test and follow-up than at pre-test.
The graph of the BDI-II results (Figure 3) showed an almost straight line in the WAT condition, therefore an exploratory analysis was done. The immediate changes in BDI-II score between pre-test and post-test, and between pre-test and follow-up were examined for the three conditions. Results unexpectedly revealed no significant effect between pre-test and post-test (t = 0.931, df = 16, p = .366), or between post-test and follow-up (t = .226, df = 16, p = .824) for the WAT condition. Furthermore, there was no significant effect between pre-test and post-test (t = 0.587, df = 13, p = .567), or between post-test and follow-up (t = 2.029, df = 11, p = .067) for the WL condition. Finally, there unexpectedly was a significant effect between the pre-test and post-test (t = 2.193, df = 16, p = .043), and between post-test and follow-up (t = 3.006, df = 15, p = .09) for the WT condition. This means that the depressive symptoms did not decrease in the WAT or WLC condition between pre-test and post-pre-test, or between post-pre-test and follow-up. Only participants in the WT condition experienced a decrease in their depressive symptoms, both between pre-test and post-test, and between post-test and follow-up. This significant effect in the WT condition explains the main-effect of time for the BDI-II score.
16 Figure 2. Repeated-measures analysis: Posttraumatic Diagnostic Scale (PDS) scores, plotted for the different conditions from pre-test to follow-up. WAT = Writing about Trauma; WT = Writing about something Trivial; WL = Waiting List.
Figure 3. Repeated-measures analysis: Beck Depression Inventory (BDI) - II scores, plotted for the different conditions from pre-test to follow-up. WAT = Writing about Trauma; WT = Writing about something Trivial; WL = waiting List.
Moderation analyses
Results showed no moderation effects regarding the difference score of the PDS between pre-test and follow-up for the PTCI score, b = 0.058, t = 1.040, p = .305, the I-E locus of control scale score, b = -0.272, t = -0.674, p = .504, or the TAS-20 score, b = 0.272, t = 1.351, p = .185. Furthermore, no moderation effects were found regarding the difference score of the BDI-II between pre-test and
14 16 18 20 22 24 26
Pre-test Post-test Follow-up
PDS -S co re Time WAT WT WL 13 14 15 16 17 18 19 20
Pre-test Post-test Follow-up
BD I-II -S co re Time WAT WT WL
17 follow-up for the PTCI score, b = 0.009, t = 0.176, p = .861, the I-E locus of control scale score, b = 0.051, t = 0.137, p = .892, or the TAS-20 score, b = 0.289, t = 1.598, p = .118. The differences in PDS-score and BDI-II PDS-score between pre-test and follow-up were not moderated by the PDS-score on the TAS-20, the PTCI or the I-E locus of control scale.
Discussion
This study examined the hypotheses that (1a-1b) a single-assignment writing therapy alleviates PTS- and depressive symptoms and (2a-2c) the effect of the assignment on the PTS- and depressive symptoms is moderated by posttraumatic cognitions, the locus of control and
alexithymia. Our findings did not support the hypotheses; participants in all groups experienced equal reductions of PTS- and depressive symptoms after a 1-month follow-up period. The reduction of the depressive symptoms was fully explained by the reduction in the WT condition. Finally, posttraumatic cognitions, the locus of control and alexithymia did not moderate the effect of the writing assignment on PTS- and depressive symptoms.
These findings may be explained in a number of ways. Written imaginal exposure normally consists of several exposures, instead of only one writing assignment (van Emmerik et al., 2013). It was already known that a single session debriefing after a psychological trauma does not improve natural recovery from trauma-related disorders (van Emmerik, Kamphuis, Hulsbosch, & Emmelkamp, 2002). However, the current study suggests that a single trauma-focused writing therapy, as opposed to a multiple assignment writing therapy, also does not result in more improvement than when one is waiting/naturally recovering. On the contrary: writing about something trivial was as effective as writing about the traumatic event in reducing PTS symptoms and only those in the WT condition experienced an improvement in their depressive symptoms. While it is understandable that there is no recovery of depressive symptoms when someone has to wait, as was the case in the WL
condition, it is unexpected that participants in the WT condition experienced a reduction of their depressive symptoms. It could be that the instruction in the WT condition not to focus on their feelings makes participants less aware of their depressive symptoms. However, since there also is a
18 significant reduction in the depressive symptoms between the post-test and follow-up this seems less likely. The lack of improvement in the WAT condition might be the result of the exposure during the assignment: writing about the trauma gives the exposure needed to improve in PTS symptoms, but the depressive symptoms remain when only one assignment is made. It might be a result worth investigating in a future study, since earlier research suggests depressive symptoms do improve after a multiple assignment writing therapy (van Emmerik, Reijntjes, & Kamphuis, 2013).
A second explanation of the results could be that an initial decrease is caused by the expectation someone has when participating in the study. Potential participants signed up with the idea of getting an intervention to alleviate their PTS symptoms which might have caused an
alleviation in their symptoms. However, the follow-up period used in this study was one month and it is highly unlikely that the expectation effect lasts this long in the waiting list group. Especially since participants in the waiting list group knew that they had to wait and therefore knew that they would not get an intervention that could reduce their symptoms.
A third explanation could be that the pre-test effect overruled any differential effects of trauma-focused writing compared to non-trauma-focused writing or waiting. Participants in all conditions completed a pre-test that required detailed disclosure of PTS symptoms and the traumatic experience, therefore possibly creating an exposure like the trauma-focused writing assignment. This would be in line with earlier research that showed an immediate decrease in PTS symptoms not only for those who wrote about their trauma, but also for those who wrote about something trivial (Truijens & van Emmerik, 2014). Even though the current study administered the writing assignment one week after the pre-test instead of immediately, like in the study of Truijens and van Emmerik (2014), it could be that the pre-test still overruled any differential effect of the PTS symptoms. Since not only the participants in the writing conditions experienced an improvement regarding the PTS symptoms, but also those in the waiting list condition, the most parsimonious explanation is that there is an effect of the pre-test on the PTS symptoms. It would be interesting to examine whether multiple writing assignments contribute to this effect on the PTS- and depressive symptoms.
19 A limitation of this study is the fact that the writing assignments were not read by the
researchers, therefore it could be possible that participants did not expose themselves enough. For example, when writing about the trauma it is important to describe the trauma in the present tense, but it might be that participants tend to write in the past tense to avoid exposing themselves. However, this is unlikely since earlier research from Truijens and van Emmerik (2014) showed that someone who has made a writing assignment experiences an immediate decrease in avoidant behaviour probably explained by the exposure. Furthermore, the lack of improvement in the WAT condition regarding the depressive symptoms seems to suggest that the participants did expose themselves more than participants in the WT condition.
A second limitation is that participants were divided in a certain order. Since it was known that participant one was assigned to condition one, the allocation was predictable and lacked randomization. Therefore, the researcher was not blind to the condition. The possibility to know this might result in more optimistic contact with participants in the WAT condition than participants in the WT or WL condition. This could result in more improvement in participants in the WAT condition. However, it is highly unlikely that this influenced the research, since all contact between the
researcher and the participants was standardized. Furthermore, results unexpectedly showed an equal decrease in PTS symptoms in all conditions and only a decrease in the depressive symptoms in the WT condition.
This study’s exploration of the possibility to intervene with a single writing assignment is also characterized by a number of strengths. Despite a high exclusion rate, this study did not change the used cut-off score to maximize generalizability to samples with a PTSD diagnosis. The writing assignment used in the experimental condition was highly similar to the writing assignment used in actual writing therapies; minimal adaptations were made to fit the present experimental setting. Therefore, it is remarkable that this clinical sample experienced a substantial reduction in PTS symptoms.
20 The current study has tried to deepen the understanding of the possibility to alleviate PTS- and depressive symptoms immediately with a single-assignment writing therapy, as opposed to multiple assignment writing therapy. In conclusion, a single-assignment writing therapy seems to be ineffective in decreasing the PTS- and depressive symptoms beyond the effect of the disclosure during the pre-test. There was an equal decrease in PTS symptoms regardless of whether someone was waiting for therapy, was writing about neutral images or was writing about a trauma. However, only those who wrote about a neutral image experienced a decrease in depressive symptoms. To deepen our understanding of the mechanisms underlying the effect of writing interventions for PTS, a future study might focus on the contribution of multiple writing assignments to the effect of the pre-test on the PTS- and depressive symptoms.
21 References
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4thed., text rev.). Arlington: American Psychiatric Press Inc.
Andriessen, J. H. T. H., & van Cadsand, J. P. (1983). Een analyse van de nederlandse i-e-schaal. Nederlands Tijdschrift voor de Psychologie, 38, 7-24.
Ashley, L., O’Connor, D. B., & Jones, F. (2011). Effects of emotional disclosure in caregivers: Moderating role of alexithymia. Stress and Health, 27, 376-387.
Blaauw, E., Voort, M., & Kerkhof, A. J. F. M. (1999). Suïcides in detentie. The Hague, the Netherlands: Elsevier.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). BDI-II, beck depression inventory: Manual (2nd ed.). San
Antonio: Psychological Corporation.
Dunmore, E., Clark, D. M., & Ehlers, A. (1999). Cognitive factors involved in the onset and
maintenance of posttraumatic stress disorder (ptsd) after physical or sexual assault. Behaviour Research and Therapy, 37, 809-829.
Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38, 319-345.
Foa, E. B., Cashman, L., Jaycox, L., Perry, K. (1997). The validation of a self-report measure of posttraumatic stress disorder: The posttraumatic diagnostic scale. Psychological assessment, 9, 445-451.
Foa, E. B., Cashman, L., Jaycox, L., & Perry, K. (1997). The validation of a self-report measure of posttraumatic stress disorder: The posttraumatic diagnostic scale. Psychological Assessment, 9, 445-451.
22 Foa, E. B., Ehlers, A., Clark, D. M., Tolin, D. F., & Orsillo, S. M. (1999). The posttraumatic cognitions
inventory (ptci): Development and validation. Psychological assessment, 11, 303-314.
Frewen, P. A., Lanius, R. A., Dozois, D. J. A., Neufeld, R. W. J., Pain, C., Hopper, J. W., Densmore, M., & Stevens, T. K. (2008). Clinical and neural correlates of alexithymia in posttraumatic stress disorder. Journal of Abnormal Psychology, 117, 171-181.
Hayes, A. F. (2013). An introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York: The Guilford Press.
Hendriks, G-J., de Kleine, R., & van Minnen, A. (2015). Optimizing the efficacy of exposure in ptsd treatment. European Journal of Psychotraumatology, 6, 1-2.
Lange, A., Schrieken, B., Blankers, M., van de Ven, J-P., & Slot, M. (2000). Constructie en validatie van de gewaarwordingenlijst (gl): Een hulpmiddel bij het signaleren van een verhoogde kans op psychosen. Directieve Therapie, 20, 82-87.
Lange, A., Schrieken, B. A. L., Scheijde, R., Broeksteeg, J., Ruwaard, J., Schrijver, M., …. Emmelkamp, P. M. G. (2005). Interapy: Diagnostiek en geprotocolleerde behandeling van welomschreven stoornissen via internet. Tijdschrift voor Psychotherapie, 31, 355-376.
Lange, A., Schrieken, B., van de Ven, J. P., Bredeweg, B., Emmelkamp, P. M. G., van der Kolk, J., … Reuvers, A. (2000). “Interapy”: The effects of a short protocolled treatment of posttraumatic stress and pathological grief through the internet. Behavioural and Cognitive Psychotherapy, 28, 175-192.
McCarthy, S. (2008). Posttraumatic stress diagnostic scale (pds). Occupational Medicine, 58, 379.
Parker, J. D., Eastabrook, J. M., Keefer, K. V., & Wood, L. M. (2010). Can alexithymia be assessed in adolescents? Psychometric properties of the 20-item toronto alexithymia scale in younger, middle, and older students. Psychological Assessment, 22, 798-808.
23 Parkers, J. D. A., Graeme, J. T., & Badby, R. M. (2003). The 20-item toronto alexithymia scale: III,
reliability and factorial validity in a community population. Journal of Psychosomatic Research, 55, 269-275.
Resick, P. A. (2000). Stress and trauma. Londen: Taylor & Francis.
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological monographs: General and Applied, 80, 1-28.
Ruwaard, J., Lange, A., Schrieken, B., Dolan, C. V., & Emmelkamp, P. (2012). The effectiveness of online cognitive behavioral treatment in routine clinical practice. PLoS One, 7, e40089.
Schnurr, P. P, Chard, K. M., Ruzek, J. I., Chow, B. K., Shih, M. C., Resick, P. A, ... Lu, Y. (2015). Design of va cooperative study #591: CERV-PTSD, comparative effectiveness research in veterans with ptsd. Contemporary Clinical Trials, 41, 75-84.
Sprinkle, S. D., Lurie, D., Insko, S. L., Atkinson, G., Jones, G. L., Logan, A. R., & Bissada, N. N. (2002). Criterion validity, severity cut scores and test-retest reliability of the beck depression inventory-II in a university counseling center sample. Journal of Counseling Psychology, 49, 381-385.
Stander, V. A., Thomsen, C. J. & Highfill-McRoy, R. M. (2014). Etiology of depression comorbidity in combat-related ptsd: A review of the literature. Clinical Psychology Review, 34, 87-98.
Strickland, B. R. (1978). Internal-external expectancies and health-related behaviors. Journal of Consulting and Clinical Psychology, 46, 1192-1211.
Taylor, G. J., Ryan, D., & Bagby, R. M. (1985). Toward the development of a new self-report alexithymia scale. Psychotherapy and Psychosomatics, 44, 191-199.
Truijens, F. L., & van Emmerik, A. A. P. (2014). Visual feedback in written imaginal exposure for posttraumatic stress: A preliminary study. Journal of Loss and Trauma, 19, 403-415.
24 Van der Oord, S., Lucassen, S., van Emmerik, A. A. P., & Emmelkamp, P. M. G. (2010). Treatment of
post-traumatic disorder in children using cognitive behavioural writing therapy. Clinical Psychology and Psychotherapy, 17, 240-249.
Van Emmerik, A. A. P., Kamphuis, J. H., & Emmelkamp, P. M. G. (2008). Treating acute stress disorder and posttraumatic stress disorder with cognitive behavioral therapy or structured writing therapy: A randomized controlled trial. Psychotherapy and Psychosomatics, 77, 93-100.
Van Emmerik, A. A. P., Kamphuis, J. H., Hulsbosch, A. M., & Emmelkamp, P. M. G. (2002). Single session debriefing after psychological trauma: A meta-analysis. Lancet, 360, 766-771.
Van Emmerik, A. A. P., Reijntjes, A., & Kamphuis, J. H. (2013). Writing therapy for posttraumatic stress: A meta-analysis. Psychotherapy and Psychosomatics, 82, 82-88.
Van Emmerik, A. A. P., Schoorl, M., Emmelkamp, P. M. G., & Kamphuis, J. H. (2006). Psychometric evaluation of the Dutch version of the posttraumatic cognitions inventory (ptci). Behavior Research and Therapy, 44, 1053-1065.
Wang, Y-P., & Gorenstein, C. (2013). Assessment of depression in medical patients: A systematic review of the utility of the beck depression inventory-II. Clinics, 68, 1274-1287.
Wohlfarth, T. D., Van den Brink, W., Winkel, F. W., & Ter Smitten, M. (2003). Screening for
posttraumatic stress disorder: An evaluation of two self-report scales among crime victims. Psychological Assessment, 15, 101-109.
